GCS_Mavlink: Make rssi adhere to MAVLink standard #3

[olliw42]

continuation of #18244

all done as suggested, rebased, cleaned up, etc

the question which scaling to choose has unfortunately still not been finally decided

for your convenience, I recall #18244 (comment), which was

concerning scaling, we talk about two scenarios, in my understanding:

scenario A:

rssi = read_receiver_rssi() * 255;
return (rssi < 255) ? rssi : 254;

scenario B:

return read_receiver_rssi() * 254;

In scenario A, the rssi values would be equal across the different telemetries (and logs?)(and support tools), except of 255 which is excluded, i.e. the range of 0 ... 254 would not exactly be 0 ... 1, but only 0 ... 0.9961

In scenario B the rssi would be 0.39% smaller than different telemetries (and logs?)(and support tools), but the range of 0 ... 254 would exactly correspond to 0 ... 1.

in this PR it is scenario A so far. Please tell/decide if that is OK or should be changed to scenario B.

[peterbarker]

LGTM